عرض عملي: كتابة ADR لمخزن بيانات المراجعات

خطوة بخطوة

1. Context: Shop expects 2 M reviews in a year, 50 writes per minute, reads by product and a rating summary, a flexible optional "attributes" part.
2. List options: PostgreSQL (with jsonb), MongoDB, wide-column.
3. Score them against the workload and operational cost; include the extra database cost from L03.
4. Planned mistake: choose MongoDB because it is "good for reviews", with no numbers. Fix: attach the volume and query evidence; note that 2 M rows with these queries fit PostgreSQL easily.
5. Write the decision, consequences, and a revisit trigger (for example "reconsider above 200 M reviews or 5 000 writes/s").

اختبر فهمك